capture one is just an instable software

[dougpeterson]

Well, "the math" is called RAW conversion

The part that fills in the missing colors of a pixel is called "demosaicing" (see http://en.wikipedia.org/wiki/Demosaicing).

"De-convolution" is a technique that undoes a "convolution" - usually a blur. Ye olde Unsharp Mask is an example and more advanced techniques can be useful (correcting for lens artifacts is one example).

Sure. But I was trying to simplify for him. If you'd like I can reply with a few pages of technical mumbo jumbo but it didn't seem appropriate for the intended audience.

 

[EsbenHR]

Sure. But I was trying to simplify for him. If you'd like I can reply with a few pages of technical mumbo jumbo but it didn't seem appropriate for the intended audience.

Yeah, I got that. I just think "deconvolution" was an inappropriate term and "demosaicing" would be neither more nor less mumbo jumbo for actual real memers of our species.

I have seen someone advertising "improved deconvolution" for CO5 compared with CO4 and not knowing what it meant. Now I assume demosaicing was the intended term.

 

[cam]

if anyone cares -- help me, please!!!!!!!!!!!!!!!!!!!!!!!!!!!!

i'm running 10.5.8, did a clean install of the newest version.... after importing images, nothing showed up... i made sure to remove all Cols and, after a few stutters, everything has been smooth sailing.... very stable.

i was given a chance today to test a pre-production camera (i.e., time critical) and so removed my previous sessions from my hard drive to make room for the new... everything says it imports but nothing appears (except for one -- seemingly a lucky fluke).

i am at my wits end!

that one image that survived (shot at 2500) has made me not want to be forced to use my other RAW processors. i need this working ASAP or i'm toast. the company is expecting them up!

thank you....

 

[Guy Mancuso]

Forget importing them just download the files direct from the CF or SD card Cam than bring them in.

 

[Jack]

Yeah, I got that. I just think "deconvolution" was an inappropriate term and "demosaicing" would be neither more nor less mumbo jumbo for actual real memers of our species.

I have seen someone advertising "improved deconvolution" for CO5 compared with CO4 and not knowing what it meant. Now I assume demosaicing was the intended term.

No, deconvolution is the term. Without getting overly technical:

Deconvolution enhances detail by applying an algorithm (a mathematical function) to the matrix of data to reconstruct original data that has been corrupted by (or more accurately "convolved" with) some other undesired data. In the case of digital imaging, that undesired data is usually noise of some sort. So the original data gets corrupted to varying degrees by noise and generates incorrect final values. The deconvolution algorithm is applied to the incorrect final value to back out the error caused by corrupter, renedering the correct value.

Demosaicing is the process that generates an accurate color from incomplete original color data generated by the Bayer filter.

So in a vast simplification, once you've demosaic'd, you have some reasonably accurate color along with varying degrees of corrupted color. You can now deconvolute that data to repair the corrupted data to more accurate and discrete pixel colors which will further enhance the image. The better your deconvolution algorithm is, the more discrete and accurate the color of each pixel is; the more discrete and accurate the color of each pixel is, the more fine detail you will be able to discern in your image.


Cheers,

 

[cam]

looking up how to do that as we speak... (past 2 a.m. and i'm a bit delirious) and need to be up early to return camera.

 

[Jack]

if anyone cares -- help me, please!!!!!!!!!!!!!!!!!!!!!!!!!!!!

i'm running 10.5.8, did a clean install of the newest version.... after importing images, nothing showed up... i made sure to remove all Cols and, after a few stutters, everything has been smooth sailing.... very stable.

i was given a chance today to test a pre-production camera (i.e., time critical) and so removed my previous sessions from my hard drive to make room for the new... everything says it imports but nothing appears (except for one -- seemingly a lucky fluke).

i am at my wits end!

that one image that survived (shot at 2500) has made me not want to be forced to use my other RAW processors. i need this working ASAP or i'm toast. the company is expecting them up!

thank you....

Look in the default import folder located at: Users>username>pictures>CaptureOneLibrary

 

[cam]

you guys must be magic because it behaved itself when i used your names... will look into your suggestion tomorrow (if i ever get sleep) to try and get to the root...

whatever -- bless you! your mojo worked!

xxx

 

[EsbenHR]

No, deconvolution is the term. Without getting overly technical:

No problem getting technical... I can follow you as far as you want to go

Deconvolution enhances detail by applying an algorithm (a mathematical function) to the matrix of data to reconstruct original data that has been corrupted by (or more accurately "convolved" with) some other undesired data. In the case of digital imaging, that undesired data is usually noise of some sort.

OK, so this is the culprit. Actually, in my technical mind "convolution" is a well defined mathematical operation. In lay terms, and in nature, we usually have convolution = blur. In engineering, convolution can also sharpen.

Demosaicing is the process that generates an accurate color from incomplete original color data generated by the Bayer filter.

Yes.

So in a vast simplification, once you've demosaic'd, you have some color along with some corrupted color, and you can then deconvolute that data to repair the corrupted data to more accurate and discrete pixel colors which will further enhance the image.

I disagree on your use of the term deconvolution. Putting the tech-talk aside, am I correct that you find that CO5 renders details better than CO4?

 

[Jack]

I disagree on your use of the term deconvolution.

You are free to disagree with my definition and use of the term deonvolution, but I just typed it into Google and one of the first entries returned was this: "Deconvolution is the process of filtering a signal to compensate for an undesired convolution."

Putting the tech-talk aside, am I correct that you find that CO5 renders details better than CO4?

Yes, that is what *I* have found for both my Phase P45+ and Canon 1Ds3 files. I cannot speak to other cameras.

 

[EsbenHR]

You are free to disagree with my definition and use of the term deonvolution, but I just typed it into Google and one of the first entries returned was this: "Deconvolution is the process of filtering a signal to compensate for an undesired convolution."

Yes - but (simplified again) convolution = blur, so deconvolution is intended to undo a blur. This could fx be an Unsharp Mask (technically only threshold=0 qualifies as a convolution).

This will, in fact, always increase noise, so we need noise reduction. To do noise reduction via convolution you need to blur. Pretty much what Lightroom used to do and what many DSLRs do when they process to JPEG

Yes, that is what *I* have found for both my Phase P45+ and Canon 1Ds3 files. I cannot speak to other cameras.

Great. I was mostly trying to figure out what you meant and I did not get your terminology being such a techhead...

 

[Henry Goh]

It behooves me that a thread about the stability of C1 can get so convoluted.....

 

[Guy Mancuso]

I'm starting to have deconvolutions myself. Someone pass me the bottle of Vodka

 

[Henry Goh]

after half a bottle of Vodka you should get a little convoluted (=blur) so that should be a good idea....

 

[Jack]

Well, this is why I try to avoid the technical discussions

Fact is, I am (actually was a long time ago) a mathematician. The term "convolve" is a very specific math term, which basically is a combination of functions. Deconvovlute is a reversal of a convolve, basically by dividing out by the convolve function to return to an original function. (And yes, I agree that the typical sharpening we do is a convolution function as it respects the original sensor data. As such, it is accurate to say it can deconvolute to some degree with certain settings, but I think it is more accurate to say the math applied during image sharpening is a convolution function acting on the original data to achieve a desired result.)

However, all this to me is somewhat irrelevant as nowadays I'm a digital photographer. And in digital imaging, deconvolute has an established and accepted meaning as it respects raw digital file processing, so that's the definition I use

 

[Bob]

Ah,
deconvolutions, just the sort of thing that I loved to get my teeth into when I was in school..
We used to use fourier methods to deconvolve the spectrum of copper kalpha and kbeta xray radiation from the back-scatter signature from a rotating lithium crystal. I wish I had a computer as powerful as my mac back then.
Anyway, when two functions are combined the result is a convolution.
If one had the function of say, an AA filter, defined as a spatial function, then is is possible, to the extent that the defined function for the AA filter is accurate, and that information in the convolved function is not lost, to actually remove it.
We have the situation now, where some AA filters produce effects more than just low pass, due to the fact that they are not just perfect low-pass filters, they have other optical properties.
The theory is that those other optical properties can be removed from the image.
Actually, in fourier space, it is not all that difficult, and there are a number of algorithms coming to the fore many using other techniques. So far, none are perfect, but that is due to the loss of information in a low pass that filters at a higher frequency than twice nyquist, and that the function describing the other optical properties are not perfectly defined for every camera.
-bob

 

[Jamesmd]

:clap: even I understood , I think .

 

[Guy Mancuso]

I'm reaching for the second bottle. LOL

 

[David K]

Ah,
deconvolutions, just the sort of thing that I loved to get my teeth into when I was in school..
We used to use fourier methods to deconvolve the spectrum of copper kalpha and kbeta xray radiation from the back-scatter signature from a rotating lithium crystal. I wish I had a computer as powerful as my mac back then.
Anyway, when two functions are combined the result is a convolution.
If one had the function of say, an AA filter, defined as a spatial function, then is is possible, to the extent that the defined function for the AA filter is accurate, and that information in the convolved function is not lost, to actually remove it.
We have the situation now, where some AA filters produce effects more than just low pass, due to the fact that they are not just perfect low-pass filters, they have other optical properties.
The theory is that those other optical properties can be removed from the image.
Actually, in fourier space, it is not all that difficult, and there are a number of algorithms coming to the fore many using other techniques. So far, none are perfect, but that is due to the loss of information in a low pass that filters at a higher frequency than twice nyquist, and that the function describing the other optical properties are not perfectly defined for every camera.
-bob

Perfect example of the lay definition of convoluted

 

[Bob]

This is piece of cake compared to just about anything that goes through congress, and I am referencing no bill in particular.
-bob